Nuclear financing in the new era

Something structural has changed in nuclear energy finance. Between 2020 and 2025, capital flows to nuclear power grew by 50 percent. They were on course to reach approximately $75 billion in 2025 alone, compared to a 2017–2023 annual average of around $50 billion and surpassed $80 billion by the end of 2025.

On June 11, 2025, the World Bank lifted its longstanding prohibition on nuclear project financing, which had been in place since 2013.  The same week, Microsoft’s 20-year power purchase agreement with Constellation Energy to restart Three Mile Island was progressing toward its 2028 completion target. Amazon’s data centres in Virginia were drawing on 1,920 megawatts of nuclear output from the Susquehanna plant. Google had signed the first corporate SMR fleet deal in history with Kairos Power.

And yet, in project development offices and government ministries across Europe, North America, and emerging nuclear markets, the approach to capital markets has not changed materially since the previous build cycle ended. The pitch decks tell a technology story. The risk allocation follows a model that has already destroyed two major developers. The financing structure assumes a type of capital that no longer dominates the market.

So what does the new nuclear capital stack actually look like? What has changed, which instruments are now available? Are developers, vendors, and governments taking the right approach when trying to attract it?

The scale of the shift

Nuclear investment has not merely recovered. It has been structurally repositioned. The IEA’s World Energy Investment 2026 report projected global nuclear capital flows to surpass $80 billion in 2025, roughly double the 2018 low, driven by new-build programmes, life-extension investments, and advanced-reactor development.

The composition of that capital has changed as much as its volume. A decade ago, nuclear financing was predominantly public: government balance sheets, state utilities, and bilateral development finance for export programmes. The private sector was largely absent. Today it is present in forms that would have seemed implausible in 2015.

Big technology companies have emerged as one of the most consequential new sources of capital in the sector. Over the twelve months to mid-2026, major technology companies signed contracts representing more than 10 gigawatts of possible new nuclear capacity in the United States. Microsoft committed to a 20-year PPA with Constellation Energy for the restart of Three Mile Island Unit 1. Amazon signed an agreement for 1,920 megawatts of output from the Susquehanna nuclear plant to power its AWS data centres and invested $500 million in X-Energy’s SMR programme. Google signed the first corporate SMR fleet deal, securing 500 megawatts from Kairos Power. Meta issued a request for proposals for one to four gigawatts of new nuclear generation and has awarded Oklo, TerraPower and Vistra – in addition to Constellation previously.

The development finance community has undergone a parallel recalibration. The World Bank’s June 2025 policy reversal was the most visible signal, but it is part of a broader pattern of multilateral institutions revisiting prohibitions that were adopted in a different energy policy environment. The Bank’s new framework explicitly supports SMR deployment, reactor life-extension programmes, and regulatory capacity building in emerging markets.

In the advanced reactor sector, private capital has accelerated sharply. Between 2020 and 2025, advanced reactor technologies secured $3.9 billion in funding. TerraPower, developing a sodium-cooled fast reactor in Wyoming, closed a $650 million Series C in June 2025, backed by Bill Gates and NVIDIA’s venture arm. In the final quarter of 2025 alone, nuclear startups raised more than $1 billion in a five-week period.

Three instruments that are reshaping the capital stack

The volume of capital moving toward nuclear is one part of the story. The more consequential change is structural: the instruments through which nuclear programmes are being financed have diversified significantly, and the logic of risk allocation underpinning them has shifted. Three instruments deserve particular attention.

The Regulated Asset Base (RAB) model. The most significant structural innovation in nuclear project finance over the past decade is the RAB model, first applied in the UK for large infrastructure projects and now being used for Sizewell C. Under the conventional Contracts for Difference (CfD) framework, the developer finances construction and begins receiving revenue only when the plant generates electricity, meaning all construction risk sits on the project company and its lenders. The cost of capital under this structure has been estimated at approximately two-thirds of the overall project cost.

The RAB model distributes risk differently: consumers contribute a regulated levy during construction, reducing the burden on the project’s debt stack and substantially lowering the cost of capital. The UK government estimates that the RAB approach for Sizewell C could save consumers £30 billion over the project’s life compared to a CfD structure, not because the plant is cheaper to build, but because it is cheaper to finance. Sizewell C reached financial close in November 2025 with a £38 billion financing package including £5 billion in export credit-backed debt from 13 banks, equity from the UK National Wealth Fund, and an investment-grade credit rating.

Export credit agency financing. Export credit agencies have become a defining feature of competitive nuclear deal structures in international markets. France’s Bpifrance provided the anchor export credit facility for Sizewell C. The US Export-Import Bank approved a $98 million loan to support a front-end engineering and design study for a US SMR export project in Romania, as part of a broader strategy to counter Russian and Chinese state-backed nuclear finance in emerging markets. South Korea’s KEXIM has been similarly active in supporting APR1400 exports, most notably through the Barakah project in the UAE. The competition among nuclear-exporting nations- the US, France, South Korea, Russia, China – is now as much a competition over financing terms as it is over technology or regulatory pathways.

Technology offtake as credit support. The most structurally novel development is the use of long-term power purchase agreements from creditworthy technology companies as the anchor offtake underpinning project finance debt. A 20-year PPA with an investment-grade counterparty provides the revenue certainty that project finance lenders require to underwrite construction risk. This mechanism, standard in renewable energy project finance for two decades, is now being applied to nuclear for the first time at scale. It is not a solution for all nuclear programmes, but it has materially changed the financing logic for both operating plant acquisitions and new-build in markets with data-centre demand.

What developers are still getting wrong

Against this backdrop of genuine structural change, the gap between what the capital market now offers and how most nuclear developers approach it remains wide. Three persistent failures stand out.

Telling a technology story instead of a commercial story. The dominant mode of presenting nuclear projects to investors and lenders remains technology-centric: reactor design, fuel cycle, safety case, regulatory status. These are necessary but not sufficient. What capital allocators need to assess is a commercial story: who is the offtaker, what is the revenue structure, how is construction risk allocated, what is the exit? Most nuclear pitch decks are engineering documents with financial appendices. The most successful recent capital raises, TerraPower’s Series C, the Amazon-Talen transaction, the Sizewell C financing, are defined by their commercial architecture, not their technology specification.

Misallocating risk across the capital stack. The failure of the fixed-price EPC model at Vogtle and V.C. Summer demonstrated the consequences of concentrating construction risk in a single contractor. The lesson has been partially absorbed: Sizewell C’s financing structure deliberately distributes risk across consumers, the state, export credit agencies, and commercial lenders. But many programmes in development, particularly in Europe and emerging markets, continue to propose structures that aggregate risk in ways that no private lender will absorb at a viable cost of capital. The result is a financing gap that cannot be bridged by optimism about future cost reductions.

Failing to differentiate in a crowded market. The expansion of the advanced reactor development pipeline has created a differentiation problem. Dozens of programmes are making broadly similar value propositions, clean, firm, dispatchable, to the same pool of institutional and venture capital. In the absence of operational data, capital has tended to consolidate around programmes with credible near-term deployment pathways, strong sponsor balance sheets, and anchor offtake. Programmes that cannot demonstrate at least two of these three characteristics are finding the capital market unreceptive, regardless of the quality of the underlying technology.

What the right approach looks like

The nuclear programmes that have successfully raised capital in the current environment share a set of structural characteristics that are instructive for developers at earlier stages.

They separate development risk, construction risk, and operational risk, and match each to the appropriate capital type. Development-stage risk, which is highest and most binary, attracts equity from sponsors, governments, and strategic investors. Construction risk, which is the primary source of cost overruns in nuclear history, is distributed across ECA facilities, regulated mechanisms, and consumer contributions where available. Operational risk, which is the most predictable phase of a nuclear project’s life, supports long-term project finance debt and institutional bond-market instruments.

They build the revenue story before they build the financing structure. The Constellation-Microsoft transaction worked because the commercial logic preceded the financing. The PPA created offtake certainty, allowing Constellation to finance the restart of Three Mile Island with a conventional debt-equity structure. Developers seeking to attract institutional debt without a credible offtake arrangement are, in the current market, asking lenders to take a risk that the technology sector has shown is transferable to creditworthy counterparties.

They engage export credit agencies as strategic partners, not as lenders of last resort. The ECA market is competitive. France, the US, South Korea, and others are actively seeking transactions that demonstrate the commercial viability of their respective nuclear export programmes. Developers with the right geographic and technological profile can use ECA engagement to anchor their financing structure, reduce their cost of capital, and signal to commercial lenders that the project’s sovereign risk dimension is managed.

Finally, they invest in owner capability alongside the capital story. A financing structure is only as credible as the organisation that will execute against it. Lenders and equity investors in the current environment are conducting rigorous assessments of owner capability, the technical teams, the programme management systems, the supply chain relationships that will determine whether the construction schedule on which the financial model depends is deliverable. The programmes that struggle to close their financing are often those in which the commercial architecture is sound, but the owner organisation cannot demonstrate it can build what it proposes to finance.

The window is open, but not indefinitely

The conditions that have brought institutional capital back to nuclear energy are a function of a specific moment: rising electricity demand from data centres and electrification, geopolitical pressure on energy security, a decade of renewable energy finance experience that has trained capital markets to underwrite infrastructure risk. These conditions may persist for a decade or more. They will not persist forever, and the competitive dynamics among nuclear-exporting nations mean that programmes that fail to close financing in the current cycle may find the next window has a different set of terms.

The capital is available. The instruments are more sophisticated than they have ever been. The remaining gap is not financial. It is the gap between how nuclear programmes have historically presented themselves to capital markets and what those markets now require.

Closing that gap is not a technical problem. It is a strategic one.